Health behavior during periods of stressful uncertainty: associations with emotions, cognitions, and expectation management

Abstract

Objective: The present study examined how cognitions and emotions characteristic of awaiting uncertain news influenced healthy (diet/exercise) and unhealthy (alcohol use) behaviors in three samples of people awaiting important news.

Design: Study 1 examined voting-eligible citizens during the month prior to learning the results of the 2016 U.S. presidential election. Study 2 examined the experience of law graduates across four months while they awaited the results of their bar exam (i.e., the licensing exam they need to pass to practice law). Study 3 examined current or recent PhD students searching for a job on the academic job market.

Results: Though the findings were somewhat mixed across studies, they generally suggest a relationship between positive emotions and health-promoting behaviors and between worry and alcohol use, with less consistent relationships between outcome expectations and health behaviors.

Conclusion: Taken together, these results offer a promising set of initial findings to understand health behavior in the context of awaiting uncertain news and provide a foundation for future investigations on the topic.

Keywords:

• Health behavior
• waiting
• uncertainty
• stress
• alcohol use
• diet & exercise

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 We chose to set our threshold at a moderate level, rather than a binge drinking level, to detect even slightly above-recommended drinking levels among our participants.

2 For healthy behavior, a quadratic growth model fit best, ▵χ² = 22.3, p < .01 (compared to a linear growth model), and the fixed effect of quadratic time was significant, t = -3.65, p = .0004. The linear fixed effect was not significant, t = 1.46, p = .15. See Sweeny and Howell (2017; Study 1) for longitudinal growth model results for worry, emotions, bracing, and hope/optimism.

3 We also conducted analyses with each health behavior separately, using MPLUS to appropriately address the categorical nature of the individual items. The conclusions are generally the same at the item level, such that all individual items show a within-subject association for positive emotion; some show a weak between-subjects association for positive emotion; and three out of five items show a between-subjects association for outcome expectations. Other individual-item associations were several within-subject associations with worry, negative emotion, and bracing (more worry/negative emotion/bracing = less healthy behavior).

4 For healthy behavior, a quadratic growth model fit best, ▵χ² = 13.9, p < .01 (compared to a linear growth model), and the fixed effect of both quadratic time, t = 4.61, p < .0001, and linear time, t = -1.89, p < .06, were significant or marginally significant. For drinking, a quadratic growth model also fit best, ▵χ² = 3.8, p = .05, and the fixed effect of quadratic time was marginally significant, t = 1.75, p = .08 (linear time: t = -1.62, p = .11). Similarly, the fixed effect of quadratic time was significant for negative emotion, t = -1.99, p = .050, and positive emotion, t = 2.67, p = .01. Although other variables did not show a quadratic pattern, we nonetheless controlled for time in all analyses to be conservative.